Master's Thesis

Droplet Impingement on the Porous Iron Oxide Layer: VOF Simulation

Final Thesis 2.44 MB

Author of thesis: Ing. Xin Cheng

Acad. year: 2025/2026

Supervisor: doc. Ing. Jan Boháček, Ph.D.

Reviewer: Ing. Jiří Hvožďa, Ph.D.

Abstract:

Droplet impact on the porous iron oxide layer is a multiphase flow problem of practical significance in spray cooling, in which the cooling spray interacts with the oxide skin formed on the surface of high-temperature metal. This paper adopts the Volume of Fluid (VOF) method realized in the open source CFD software OpenFOAM, combined with the MULES algorithm for interface capture and continuous surface force model for surface tension calculation, and conducts a numerical study of this phenomenon. All solid surfaces have a static contact angle of 20°, which is consistent with the hydrophilicity of iron oxide.

The porous oxide layer is represented by a pore-resolved computational mesh instead of a volume-averaged continuum model. The grid is reconstructed from the micrograph stack of the actual oxide sample: the micrograph image is loaded into ParaView, resampled to the uniform OpenFOAM grid through the ResampleWithDataset function, denoising with a median filter, and the solid and pore regions are separated by applying an intensity threshold. The generated oxide geometry is embedded in the refined background grid through the mergeMeshes and stitchMesh functions, thus forming a single composite domain.

A reference case of droplet diameter D0 = 0.1 mm and impact speed U0 = 1 m/s was established and extended it through two parameters: one is the speed parameter, U0 is 1, 2, 5 and 10 m/s respectively; the other is the droplet size parameter D0, with values of 0.1, 0.3 and 1 mm, the latter of which is obtained by rescaling the geometry using the transformPoints function. Pore penetration is quantified by two indicators extracted along the internal reference path of the oxide: the maximum penetration depth d(t) and the interface position through specific path dint (t) that changes with time. These two indicators together reveal that the early inertial dominant stage is followed by a slower capillary-driven phase. The grid of the reference case was refined to verify the sensitivity of the grid, the current non-convergence is due to several factors, which are discussed below.

Keywords:

Volume of Fluid (VOF), Droplet impingement, Porous iron-oxide layer, OpenFOAM

Date of defence

17.06.2026

Result of the defence

Defended (thesis was successfully defended)

znamkaDznamka

Grading

D

Process of defence

The student presented the committee with the progress of the work, the results and conclusions of the thesis, and answered the supervisor's and reviewer's questions. During the defence, the student was unable to fully defend the work and did not demonstrate command of the details of the simulation; although the work employed an advanced simulation, the student was not fully familiar with its underlying details. Doc. Lízal asked about the experimental validation of the student's simulations. The student stated that no experimental work had been carried out. When asked which experimental methods could at least be used for some form of validation, the student was unable to identify any suitable method. Prof. Novotný asked whether the model was a 2D model or a genuinely axisymmetric one, and since the student had described it as axisymmetric, where the axis was located. The student indicated the position of the axis, after which the committee discussed the boundary conditions and the simulation setup. Prof. Novotný further asked which turbulence models the student had used. The student had not considered turbulence and had modelled laminar flow only. The committee discussed the suitability of this approach for the given simulation. Doc. Rudolf commented on the work and its conclusions and asked whether the velocities used were relevant to real applications. The student responded that 5 m/s was realistic. In response to a further question on whether a contact angle of 20° was realistic, the student noted that 20° could be realistic if the surface were heated. Doc. Jan asked about the physical background, namely, what happens when a liquid enters a very small, very hot pore, and how rapid the evaporation of the liquid would be. The student assumed that evaporation would be slow. Doc. Jan then asked whether the time required for the liquid to evaporate could be estimated by a simple calculation, and whether it would be on the order of milliseconds or hundreds of milliseconds. The student was unable to answer; even when guided by Doc. Jan towards thermomechanics, the student drew a T–s diagram but could not arrive at a solution. Doc. Lízal commented that the calorimetric equation might be suitable here.

Language of thesis

English

Faculty

Fakulta strojního inženýrství

Department

Heat Transfer and Fluid Flow Laboratory

Study programme

Mechanical Engineering (N-ENG-A)

Composition of Committee

doc. Ing. Pavel Charvát, Ph.D. (předseda)
prof. Ing. Pavel Novotný, Ph.D. (místopředseda)
doc. Ing. František Lízal, Ph.D. (člen)
doc. Ing. Vít Jan, Ph.D. (člen)
doc. Ing. Jiří Šremr, Ph.D. (člen)
doc. Ing. Pavel Rudolf, Ph.D. (člen)

Supervisor’s report
doc. Ing. Jan Boháček, Ph.D.

The master’s thesis by Xin Cheng, entitled Droplet Impingement on the Porous Iron Oxide Layer: VOF Simulation, addresses a demanding CFD problem involving two-phase flow, surface tension, wetting, and liquid penetration into a pore-resolved oxide structure. The student implemented the simulations in OpenFOAM using the VOF approach and developed a complex workflow for converting micrograph-based oxide geometry into a computational mesh. The work includes a reference droplet-impact case, parameter studies of contact angle, impact velocity and droplet diameter, as well as a grid-sensitivity analysis.

The main strength of the thesis is the successful construction of a complete numerical workflow combining image-based porous-structure reconstruction, OpenFOAM mesh generation, and multiphase VOF simulations. The student demonstrated independence in learning and applying advanced open-source CFD tools. The achieved results provide useful qualitative insight into the role of wettability, inertia and pore geometry in droplet penetration into an iron-oxide layer.

In relation to the interpretation of the pore-penetration mechanism, I would like the student to address during the defence how droplet permeation would change if the porous medium were open at the bottom instead of being confined, and whether the internal pore structure could theoretically prevent permeation completely even for the hydrophilic contact angle of 20°. The submitted work fulfils the main objectives of the assignment and represents a valuable first step toward pore-resolved modelling of droplet interaction with oxide layers.

I recommend the thesis for defence.
Evaluation criteria Grade
Fulfilment of requirements and objectives of assignment A
Working process, extent and suitability of applied methods B
Scholarly contribution and originality B
Ability to interpret achieved results and draw conclusions A
Applicability of results in practice or theory C
Logical arrangement of thesis and its layout A
Grafic layout, used style and language level A
Work with used sources including quotations B
Student's independence when working on the topic A

Grade proposed by supervisor: A

Reviewer’s report
Ing. Jiří Hvožďa, Ph.D.

The thesis deals with a demanding and practically relevant topic: the numerical simulation of droplet impact on a porous iron oxide layer using the VOF method in OpenFOAM. The strongest part of the work is the numerical and methodological procedure. The student developed a workflow for converting micrograph data into a pore-resolved computational mesh and used it for simulations of liquid penetration into a complex porous structure. This demonstrates good ability to work with CFD tools, mesh generation, image-based geometry reconstruction and post-processing.

The objectives of the assignment were fulfilled to a good extent. The flat-surface impact case is not explicitly presented in detail, but since the thesis successfully treats a more complex pore-resolved geometry, this can be considered only a minor shortcoming. The achieved results are physically plausible and are evaluated using suitable quantitative indicators, especially maximum penetration depth and penetration along a selected reference path. The influence of contact angle, initial velocity and droplet diameter is discussed reasonably. The interpretation would, however, be stronger if supported by experimental validation or comparison with a simpler benchmark case. Since the contact angle is identified as an important parameter, the choice of only two investigated values should also be better justified.

The main weaknesses are formal and editorial. The text contains language inaccuracies, inconsistent formatting of equations and graphs, occasional alignment problems and missing spaces. Some code is presented as screenshots, while typeset code would be preferable. Fig. 4.15 is not fully clear. Minor issues also include non-functional hyperlinks, duplicated definition of the Ohnesorge number, an unresolved reference error on page 55, unsuitable ordering of references and uncommented appendix code.

Despite these shortcomings, the thesis is concise, original and technically valuable. I recommend it for defence.
Evaluation criteria Grade
Fulfilment of requirements and objectives of assignment B
Working process, extent and suitability of applied methods A
Scholarly contribution and originality A
Ability to interpret achieved results and draw conclusions B
Applicability of results in practice or theory A
Logical arrangement of thesis and its layout C
Grafic layout, used style and language level C
Work with used sources including quotations B
Topics for thesis defence:
  1. One objective of the assignment was to simulate droplet impact on a flat surface. This case is not explicitly presented in the thesis. Was the flat-surface case performed during the development of the model? What conclusions were obtained from it?
  2. How was gravitational acceleration treated in the simulations, and how does this affect the interpretation of the prescribed velocity U0?
  3. How was the selected reference path chosen, and how sensitive are the reported penetration results to the position of this path within the porous structure?
  4. The contact angle was identified as a key parameter affecting liquid penetration into the porous oxide layer. What was the rationale behind choosing the two investigated values of 20° and 70°, and what behaviour would you expect for intermediate contact angles?

Grade proposed by reviewer: B

Responsibility: Mgr. et Mgr. Hana Odstrčilová